A portfolio management of a small RES utility with a Structural Vector Autoregressive model of German electricity markets

TL;DR

This paper develops a dynamic portfolio management strategy for small renewable energy utilities in Germany using a Structural Vector Autoregressive model to optimize trading across different electricity markets and reduce risk.

Contribution

It introduces a data-driven approach employing SVAR models for dynamic portfolio optimization in electricity trading, enhancing revenue and risk management.

Findings

Data-driven strategies increase utility revenue.

Risk is reduced through improved predictability.

Sharp Ratio-based approach yields robust results.

Abstract

The changes in electricity markets expose RES producers and electricity traders to various risks, among which the price and the volume risk play a very important role. In this research, a portfolio building strategies are presented, which allow to dynamically choose a proportion of electricity traded in different electricity markets (day-ahead and intraday) and hence to optimize the behavior of an utility. Two types of approaches are considered: simple, assuming that the proportions are fixed, and data driven, which allows for thier fluctuation. In order to explore the market information, Structural Vector Autoregressive (SVAR) model is applied, which allows to estimate the relationship between variables of interest and to simulate their future distribution. The presented methods are evaluated with data coming from German electricity market. The results indicate that data driven trading strategies allow to increase the utility revenue and at the same time reduce the trading risk, measured by the predictability of the next day income and the revenue Value at Risk. It turns out that the approach based on Sharp Ratio provides the most robust results.